Background Interpersonal psychotherapy (IPT) and cognitive behavioral therapy (CBT) are both evidence‐based treatments for major depressive disorder (MDD). Several head‐to‐head comparisons have been made, mostly in the United States. In this trial, we compared the two treatments in a small‐town outpatient psychiatric clinic in Sweden. The patients had failed previous primary care treatment and had extensive Axis‐II comorbidity. Outcome measures were reduction of depressive symptoms and attrition rate. Methods Ninety‐six psychiatric patients with MDD (DSM‐IV) were randomized to 14 sessions of CBT (n = 48) or IPT (n = 48). A noninferiority design was used with the hypothesis that IPT would be noninferior to CBT. A three‐point difference on the Beck Depression Inventory‐II (BDI‐II) was used as noninferiority margin. Results IPT passed the noninferiority test. In the ITT group, 53.5% (23/43) of the IPT patients and 51.0% (24/47) of the CBT patients were reliably improved, and 20.9% (9/43) and 19.1% (9/47), respectively, were recovered (last BDI score <10). The dropout rate was significantly higher in CBT (40%; 19/47) compared to IPT (19%; 8/43). Statistically controlling for antidepressant medication use did not change the results. Conclusions IPT was noninferior to CBT in a sample of depressed psychiatric patients in a community‐based outpatient clinic. CBT had significantly more dropouts than IPT, indicating that CBT may be experienced as too demanding. Since about half the patients did not recover, there is a need for further treatment development for these patients. The study should be considered an effectiveness trial, with strong external validity but some limitations in internal validity.